Compute similarity between columns in a fast manner MATLAB -
given (rather) big matrix :
[m*n] (m = 7, n = 15000)
where columns items , rows attributes, compute similarity between each 2 items , store in array. similarity matrix sim
each row contains [item1_id,item2_id, similarity]
. process needs done fast.
i considering options @bsxfun uses multi-threading purpose. others ideas appreciated (i totally open other efficient approaches). appreciated if suggest approaches , report timing takes perform operation. process may scaled in future.
thank much.
Comments
Post a Comment